Research Engineer, Applied AI
Agentio · New York, NY · 3 wk ago
On-siteEngineeringFull-time
We need to output HTML fragment with sections: About the role, Responsibilities, Requirements, Qualifications, Skills, Benefits, Pay, Schedule. Only include sections that have real specific content. The posting includes:
- About Agentio (description) maybe "About the role"? Actually "About Agentio" is description of company. Could be "About the role"? The instruction: use for section headings (e.g. About the role, Responsibilities, Requirements, Qualifications, Skills, Benefits, Pay, Schedule). So we need to map content to those headings.
We have:
- About Agentio: description of company and mission. Could be placed under "About the role"? Usually "About the role" is description of the job. But we have "What You'll Work On", "What You'll Do", "Who We're Looking For", "What You'll Get". So we can map:
- About the role: maybe combine the initial paragraph about Agentio and the role? But we have separate sections: "What You'll Work On" (responsibilities), "What You'll Do" (more responsibilities), "Who We're Looking For" (requirements/qualifications/skills), "What You'll Get" (benefits). Also there is a Benefits & Perks list.
Pay and Schedule: not mentioned. So we should NOT include Pay or Schedule sections because no specific content.
We need to avoid boilerplate like EEO etc. Remove that.
We need to preserve all substantive information.
Thus sections:
headings and
for section headings (e.g. About the role, Responsibilities, Requirements, Qualifications, Skills, Benefits, Pay, Schedule). So we need to map content to those headings.
We have:
- About Agentio: description of company and mission. Could be placed under "About the role"? Usually "About the role" is description of the job. But we have "What You'll Work On", "What You'll Do", "Who We're Looking For", "What You'll Get". So we can map:
- About the role: maybe combine the initial paragraph about Agentio and the role? But we have separate sections: "What You'll Work On" (responsibilities), "What You'll Do" (more responsibilities), "Who We're Looking For" (requirements/qualifications/skills), "What You'll Get" (benefits). Also there is a Benefits & Perks list.
Pay and Schedule: not mentioned. So we should NOT include Pay or Schedule sections because no specific content.
We need to avoid boilerplate like EEO etc. Remove that.
We need to preserve all substantive information.
Thus sections:
About the role
maybe include description of Agentio and the role? Could be the first paragraph: "About Agentio Agentio builds AI-native infrastructure for creator advertising..." That's about company. Could be considered "About the role". We'll put that.
Responsibilities
combine "What You'll Work On" and "What You'll Do". Could be a list.
Requirements
maybe "Who We're Looking For" includes qualifications, skills, etc. Could split into Requirements and Qualifications and Skills. But we can combine into Requirements (or Qualifications). The instruction: prefer few substantive sections over many thin ones. So we could have:
- About the role
- Responsibilities
- Qualifications (or Requirements)
- Benefits
We need to decide.
We have "Who We're Looking For": includes research judgment, demonstrated experience, PhD useful, depth in areas, ability to turn ambiguous problems, strong engineering skills, strong knowledge of current AI research. That's qualifications/skills.
We have "What You'll Get": benefits list.
Thus sections:
About the role
... description of Agentio and the role.
Responsibilities
- ... from What You'll Work On and What You'll Do
Qualifications
- ... from Who We're Looking For
Benefits
- ... from Benefits & Perks
headings and - for lists,
for prose.
We need to decide which content goes where.
Option:
About the role
About Agentio Agentio builds AI-native infrastructure for creator advertising. Our platform uses AI to turn creative and performance signals into campaign strategies, match brands with creators, recommend sponsorship prices, and review videos against brand requirements. Each campaign generates feedback: how brands respond to our recommendations, which offers creators accept, and how the resulting content performs. We build self improving loops that learn from these behaviors and outcomes to better understand preferences, predict performance, and improve campaign decisions. We're looking for a Research Engineer to advance that work through applied AI and ML research, developing models that improve the platform's predictions and automate more of the decisions behind creator advertising. The work spans research and production. You'll evaluate methods from academia and industry, adapt and train models on Agentio's proprietary data, develop new approaches when existing methods fall short, and work with engineering and product to ship them.
Responsibilities
- Multimodal intelligence. Build systems that understand creators, brands, briefs, and content across video, audio, image, and text.
- Ranking, recommendation, and prediction. Develop models that predict creator-brand fit and campaign outcomes and make better decisions from sparse, heterogeneous feedback.
- Foundation models and agents. Apply LLMs and multimodal models to campaign planning, discovery, creative understanding, execution, and optimization. Use Agentio's proprietary data for prompting, retrieval, fine-tuning, post-training, and, where it makes sense, specialized model training.
- Learning and decision-making. Turn campaign data into training signals, evals, and learning loops while tackling problems in optimization, exploration, measurement, and marketplace dynamics.
- Part of the role is deciding which advances in AI and ML are useful for Agentio and where new methods are needed.
- Own research problems from problem definition and experimentation through production.
- Evaluate relevant AI and ML methods against Agentio's data, build rigorous experiments and evaluations, and measure performance in production.
- Develop models, algorithms, training approaches, or datasets when existing techniques aren't sufficient.
- Work with engineering and product to ship successful approaches.
- Help set Agentio's research agenda and technical standards for applied ML.
Qualifications
- We're looking for someone with strong research judgment who also wants to build production systems.
- Demonstrated experience conducting applied ML research, with original work shown through production ML systems, publications, open-source research, or comparable contributions.
- A PhD or advanced degree in ML, AI, computer science, statistics, or a related field is useful but not required.
- Depth in one or more areas such as foundation models, multimodal learning, recommendation and ranking, representation learning, reinforcement learning and decision-making, causal inference, optimization, or adjacent fields.
- Ability to turn ambiguous product problems into tractable research questions, run rigorous experiments, and translate successful approaches into production systems.
- Strong engineering skills and experience taking ML work beyond prototypes.
- Strong knowledge of current AI research and the judgment to determine which methods are useful in practice.
Benefits
- Flexible PTO – Take the time you need to recharge
- Comprehensive Health Coverage – Top medical plan through Aetna
- Dental & Vision Plans – Protect your smile and your sight with coverage from Guardian
- 401(k) Retirement Plan – Invest in your future while you build your career
- $100 Monthly Lifestyle Credit – Use it to shop with our amazing Agentio Brand Partners
- Daily lunches and snacks in the office to keep you at your best
- Dinner provided Monday - Thursday
- Fully subsidized gold tier access on Wellhub (Includes LifeTime, Barry's, Solidcore, Headspace, etc)
- Gym in the building! – Your workout, right downstairs
We should not include Pay or Schedule sections.
Check that we didn't include any boilerplate like EEO. We removed that.
Check that we didn't repeat job title as heading. Not present.
Check that we didn't include any
for prose. We need to decide which content goes where. Option:
About the role
About Agentio Agentio builds AI-native infrastructure for creator advertising. Our platform uses AI to turn creative and performance signals into campaign strategies, match brands with creators, recommend sponsorship prices, and review videos against brand requirements. Each campaign generates feedback: how brands respond to our recommendations, which offers creators accept, and how the resulting content performs. We build self improving loops that learn from these behaviors and outcomes to better understand preferences, predict performance, and improve campaign decisions. We're looking for a Research Engineer to advance that work through applied AI and ML research, developing models that improve the platform's predictions and automate more of the decisions behind creator advertising. The work spans research and production. You'll evaluate methods from academia and industry, adapt and train models on Agentio's proprietary data, develop new approaches when existing methods fall short, and work with engineering and product to ship them.
Responsibilities
- Multimodal intelligence. Build systems that understand creators, brands, briefs, and content across video, audio, image, and text.
- Ranking, recommendation, and prediction. Develop models that predict creator-brand fit and campaign outcomes and make better decisions from sparse, heterogeneous feedback.
- Foundation models and agents. Apply LLMs and multimodal models to campaign planning, discovery, creative understanding, execution, and optimization. Use Agentio's proprietary data for prompting, retrieval, fine-tuning, post-training, and, where it makes sense, specialized model training.
- Learning and decision-making. Turn campaign data into training signals, evals, and learning loops while tackling problems in optimization, exploration, measurement, and marketplace dynamics.
- Part of the role is deciding which advances in AI and ML are useful for Agentio and where new methods are needed.
- Own research problems from problem definition and experimentation through production.
- Evaluate relevant AI and ML methods against Agentio's data, build rigorous experiments and evaluations, and measure performance in production.
- Develop models, algorithms, training approaches, or datasets when existing techniques aren't sufficient.
- Work with engineering and product to ship successful approaches.
- Help set Agentio's research agenda and technical standards for applied ML.
Qualifications
- We're looking for someone with strong research judgment who also wants to build production systems.
- Demonstrated experience conducting applied ML research, with original work shown through production ML systems, publications, open-source research, or comparable contributions.
- A PhD or advanced degree in ML, AI, computer science, statistics, or a related field is useful but not required.
- Depth in one or more areas such as foundation models, multimodal learning, recommendation and ranking, representation learning, reinforcement learning and decision-making, causal inference, optimization, or adjacent fields.
- Ability to turn ambiguous product problems into tractable research questions, run rigorous experiments, and translate successful approaches into production systems.
- Strong engineering skills and experience taking ML work beyond prototypes.
- Strong knowledge of current AI research and the judgment to determine which methods are useful in practice.
Benefits
- Flexible PTO – Take the time you need to recharge
- Comprehensive Health Coverage – Top medical plan through Aetna
- Dental & Vision Plans – Protect your smile and your sight with coverage from Guardian
- 401(k) Retirement Plan – Invest in your future while you build your career
- $100 Monthly Lifestyle Credit – Use it to shop with our amazing Agentio Brand Partners
- Daily lunches and snacks in the office to keep you at your best
- Dinner provided Monday - Thursday
- Fully subsidized gold tier access on Wellhub (Includes LifeTime, Barry's, Solidcore, Headspace, etc)
- Gym in the building! – Your workout, right downstairs